Hybrid ACO-PSO Drone Test Expert

The hybrid Ant-Colony/Particle-Swarm path-planning module for my GPS-less multirotor is already coded; the next milestone is to refine that C++ implementation inside ArduPilot, flash it to the airframe, and fly the remaining 50 % of our validation tests. The craft uses external ranging and inertial sensors instead of GNSS or cameras, so tight software-hardware integration—and a solid grasp of Mission Planner’s logging and parameter tools—will be essential. You will start by walking through the existing source, spotting logic or timing issues that keep the algorithm from talking cleanly to the flight stack. Once the code compiles cleanly against the current ArduCopter build, you’ll load it onto the flight controller, run bench tests, then accompany (or remotely guide) the final indoor/outdoor flights needed to verify stable hover, corridor following, and obstacle-aware rerouting. Deliverables • Corrected C++ source and compiled firmware • Mission Planner parameter file and repeatable setup steps • Flight-test plan covering pre-flight checks, sensor calibration, and safety gates • Flight logs with short analysis confirming that hybrid ACO-PSO decisions are executed in real time • Concise report summarizing results and any remaining tweaks Acceptance criteria • Vehicle completes each leg of the test plan without GPS or vision aids, maintaining prescribed corridors and avoiding obstacles • Logged paths match algorithmic waypoints within the tolerances defined in the plan • No hard faults, watchdog resets, or unexplained mode changes during flights Deep knowledge of ArduPilot internals, Mission Planner, hybrid ACO-PSO theory, and hands-on drone testing is non-negotiable, and I need someone comfortable writing and debugging code directly—rather than leaning entirely on generative tools. If you can pick up the existing work and get us through the final flights, let’s talk timelines and access to the hardware. Should finish the drone test properly along with me